Font Size: a A A

Study On The Topological Properties Of The Whole Brain Functional Connectivity Network Of Schizophrenia Based On FMRI Data

Posted on:2018-01-26Degree:MasterType:Thesis
Country:ChinaCandidate:S W YiFull Text:PDF
GTID:2334330515468312Subject:Probability theory and mathematical statistics
Abstract/Summary:PDF Full Text Request
People’s brain has 1011 neurons,neurons through the synaptic junction,these neurons have formed a highly complex network.It is of great importance to figure out the information expressed in the network,which will contribute to the resolution of the prevention and diagnosis of cognitive disorder of human beings and provide meaningful references to the development of artificial intelligence.Recently,it gradually become popular that people use graph theory to mining of human brain network info.The study of human brain network has made a series of great achievements.Firstly,we use functional magnetic resonance imaging(functional magnetic resonance imaging,fMRI)data of schizophrenia and normal control subject to construct whole brain functional connectivity network,respectively based on subject’s network level and group network level.Then,we calculate these network topology.Finally,we used statistical methods and directly drawing comparison method to explore the differences of topological properties of schizophrenia and normal control subject’s whole brain functional connectivity network.The results showed that judging from the network whole properties of the subjects,volatility of schizophrenic patients’ average degree of network、network fault tolerance、edge betweenness of network and node betweenness of network bigger than normal control subjects’.And it reflects that compared to the normal human brain network’s the information flow efficiency of the patient is lower,prone to congestion,and the network’s resistance to failure will be worse.In addition to looking for differences in network whole properties,in order to locate the specific lesion area,we also explore the differences in properties of the single node and single side,and we find that there are significant differences between schizophrenic and normal control subject in terms of left middle frontal gyrus area’s node degree、left median cingulate and paracingulate gyri area’s fault tolerance、right median cingulate and paracingulate gyri area’s clustering coefficient with its fault tolerance,left median cingulate、left median cingulate and paracingulate gyri、right median cingulate and paracingulate gyri are located to the frontal lobe,so the frontal lobe is schizophrenic central lesion area.From perspective of group network,we can explore the differences of topological property between schizophrenia group network and normal control group network by comparing direct drawing.The result is that average degree of schizophrenia group network is less than average degree of normal control group network,it shows that the network connection between each node of the patient’s brain is less than connection between normal brain network of each node,and the differences of the shortest path length also indirectly illustrates this opinion,these results also reflect the efficiency of information flow within network of patient brain is lower.The discovery of these differences in the network topological properties has provided new clues for the further study on the pathological mechanism of schizophrenia.
Keywords/Search Tags:fMRI, schizophrenia, whole brain functional connectivity network, network topology, group network
PDF Full Text Request
Related items